Please use https://gpt.dongnv.dev instead. repo: https:/nullmastermind/chatgpt-web
+ I am very interested in this repository, so you don't need to worry if there
+ hasn't been an update for a while. Simply put, it is still functioning stably
+ without the need for updates.
If you need an additional bot on Telegram, you may like this repository: https:/karfly/chatgpt_telegram_bot
Discord BOT: NullGPT#0657 https://discord.gg/kCwPAGj9Rc
- Low latency replies
- No request limits
- Message streaming
- Code highlighting
- Special chat modes. You can easily create your own chat modes by editing
config.py
- The program is lightweight and compact, simply a Python script without any accompanying database or other components.
- The software has a high level of customization, allowing you to adjust the "temperature" or use a quick selection button for "temperature" if you are not satisfied with the result.
- Use the "history: number_of_last_messages" option to utilize chat history only when necessary.
- To run a local test, simply create a
.env
file and copy the content of the.env.example
file over. Please note that you need to fill in the API KEY completely.python main.py
- To run on Docker, you need to modify the environment variables in the
docker-compose.yml
file and then rundocker-compose up --build -d
.
- Special chat modes
- Integrate the features of Langchain (https:/hwchase17/langchain) to produce more complex results.
- /code
- /assistant
- /english_translator
- /english_translator_technical
- /english_teacher
- /text_improver
- /show_history <number_of_last_messages>
- /regenerate: Regenerate the answer with customizable options that can be modified
- /estimate: Task estimate assistant
- /midjourney: Expand your midjourney ideas
You can easily add or delete as desired in config.py
- continue_conv <True/False>: Whether to continue the previous conversation
- prompt: required
- temperature: The temperature to use for message generation. Default in config.py
- history: default 0. If this option is used, it will disable the 'continue_conv' option.
- max_tokens: default 1000